Capturing satisfactory images can sometimes be a challenge under low light conditions and/or instances where high dynamic range imaging is desired. Some people will attempt to address the low light condition/high dynamic range imaging by managing the exposure time, whereby extending the exposure time can allow more light relative to the image to ultimately be captured. However motion including motion of the image capture circuitry, as well as motion within the image can be more pronounced as the exposure time is extended. Motion of the image capture circuitry can also be more pronounced in instances where the image capture circuitry is incorporated as part of a hand-held device, where the hand, which is holding the device during image capture, will often have some degree of movement (i.e. jitter) that can have an effect during a time period of image capture that has been extended to allow for the capture of more light. While increasing the exposure time can have an adverse effect on image blurring, decreasing the exposure time in low light conditions can in at least some instances result in images that are dark and noisy.
Various post image capture processing has been used to try and alleviate the undesirable aspects associated with adjusting the exposure time. In at least some instances, the results of the processing can be enhanced with a good understanding of the motion that has occurred during the exposure time associated with the image that has been captured. In some instances, motion of the image capture circuitry can be determined through the use of a separate sensor, such as a gyroscope. However, it can sometimes be difficult to synchronize the gyroscope sensor data with the time period during which the image is being captured, and/or image data is being read out of the sensor cells of an image sensor. In other instances, a corresponding motion may be derived through an analysis of the image being captured including a sequence of images. However, finding feature points within an image that can be used to track motion can sometimes be difficult. Furthermore, tracking motion via feature points can also be complicated by the fact that the feature point may be a victim of the same blurring due to motion that is affecting the overall image.
The present inventors have recognized that, the determination of a motion estimate can be improved if one or more portions of the image sensor could be configured to be sampled at a higher rate, while the remaining portion of the image sensor could be configured to be sampled at the relatively lower rate to improve aspects of the image quality during low light conditions. The determined motion estimate could then be used to address aspects of the image quality that might be negatively affected by a lower sampling rate and/or a longer exposure period, such as concerns with blurring and/or noise.